Optimize Your Analysis with the Latest Analysis Workspace Enhancements
by Ben Gaines
posted on 01-21-2016
Now that the hectic holiday season is behind us, many of us can turn our attention to optimizing our analysis efforts in 2016. Becoming more efficient with your ability to produce and share meaningful insights is one of the fastest and easiest ways to make a big difference in your organization. With the update to Analysis Workspace in Adobe Analytics today, we’ve added a handful of exciting new features and improvements that will make you and your stakeholders more efficient in your analysis and reporting. In this post, I’ll briefly take you through each of these new features which make Analysis Workspace — already the preferred analysis tool for tens of thousands of analysts — even more powerful.
Undo Keyboard Shortcut
Because Analysis Workspace is all about flexibility and ad hoc, deep-dive analysis, sometimes you add a breakdown or do a comparison that isn’t quite what you wanted, or perhaps led you to a dead end in your analysis. With the new “Undo” feature, it’s easy to roll back your changes to a previous step in your analysis journey. Undo, if used repeatedly, will sequentially remove each action you’ve taken, in order, going back to the last time you saved the project you’re working in.
You can undo a change either using your natural Undo keyboard shortcut (Cmd+Z/Ctrl+Z), or by selecting Undo from the action bar at the top of the page.
This has been one of the top two or three feature requests since we launched Analysis Workspace late last year, and we’re very excited to get it into your hands.
New Layout Options
When we built Analysis Workspace, we wanted to give you as much control as possible over the layout of your analysis projects. The theory was that the more control you have over how visualizations and data tables are placed, sized and arranged, the better you can craft a story with your data. To this end, you have always been able to resize any element in your project to 100%, 75%, 50%, or 25% of the width of the page. (You can also adjust the height of each object.) With this latest release, you can now resize and rearrange in more granular grid, designed to allow (for example) 66%/33% splits, like those shown here:
(For your information, there are actually 12 “breakpoints” as you’re dragging and resizing. The smallest size you can have for a visualization or data table, therefore, is 1/12th, or 8%, of the width of the project. That’s probably smaller than you would ever want, but for some visualiation types, like the “summary” visualizations shown above, 2/12ths, or 17%, might make sense, with some 83%-width visualizations next to them.) We see this as a great way to give you even more control over how you and your colleagues consume the insights found in your analysis projects, without sacrificing ease of setup.
As with the layout options discussed above, the introduction of conditional formatting in Freeform Analysis data tables is about helping you better tell a story with data. You’ve likely used conditional formatting in Excel or similar tools, and now it is available in Analysis Workspace to cause insights to leap off the page. When selected, conditional formatting will shade the individual cells in a table on a spectrum from green to red, making it clear which items in your table are performing above the mean and which are lagging below the mean.
You can enable conditional formatting for any column of data by clicking on the settings icon for the column. You also have the ability to let Analysis Workspace figure out what is average, or you can manually adjust the thresholds for high, average and low performance. In the example shown above, I can easily see that while my conversion rates are all pretty well centered around a mean; none particularly stands out. But in the Revenue category I’ve got one product that is killing it, and a bunch of others that are well below that midpoint. (NOTE: You could, and probably should, make a strong case for using median instead of mean for Revenue, and I could essentially do that by adjusting the upper limit, midpoint, and lower limit manually, but I did not do that in the screen shot above.)
While bullet graphs are not the only new visualization type that we’ve added to Analysis Workspace in this release, they are perhaps the most unique; as I have discussed them with analysts, few have ever heard of them. Bullet graphs were originally proposed by data visualization guru Stephen Few, and in his book, Information Dashboard Design, he challenges software vendors to adopt them. We took his challenge!
Bullet graphs look a lot like bar graphs, but they are designed to show progress toward goals, or placement along a spectrum. You can think of them like metric gauges, but without the clutter and distraction. If you’re like most organizations, you have objectives that you are measured against, and you want those to be omnipresent as you are analyzing and reporting on what’s working and what isn’t working. Bullet graphs allow you to keep those goals at the forefront of your analysis. In the screen shot above, I’m about a third of the way toward my high-end goal, and almost halfway to my low-end goal. As I accrue more revenue, the green bar will progress toward the right.
Here is how it works: You add a bullet graph to your Analysis Workspace project, and you select the table cell whose progress you want to track (think “Current Quarter Revenue” or “Video Views This Week”). Then, in the configuration settings for the bullet graph, you indicate the goals or targets that you have for that metric (e.g., “$200,000” or “60,000 views”).
Analysis Workspace then tracks progress toward those goals each time you or a colleague loads up the project. Importantly, bullet graphs in Analysis Workspace make itreally easy for anyone to visually interpret how well his or her efforts are progressing toward the goal, whether you’re a seasoned analyst, or just an occasional user of data and analysis.
Scatterplots and Treemaps
Also falling under the “new visualizations in Analysis Workspace” category, scatter plots and tree maps both provide unique new ways to display your data and glean insights visually. As with all visualizations in Analysis Workspace, these can be dragged-and-dropped from the left panel onto any project. However, you do need to have the right configuration of data elements in your data tables in order for these visualizations to work correctly.
Treemaps are perhaps the easier of these two visualizations to work with; the requirements for the data table when building a treemap are the same as when building a bar chart. You can throw just about any data table at a treemap and Analysis Workspace will create the boxes for you. As shown above, they can handle multiple metrics at once.
Scatterplots are (in my opinion) a little more useful than treemaps, but a bit more specific in terms of the data they require. The table providing the data should have at least two metric columns; the first column defines the x-axis, and the second column defines the y-axis. If there is a third metric column in your table, the scatterplot will use the third metric to determine the radius/width of the dot on the graph.
Deep Linking for Projects
Another common feature request we’ve gotten for Analysis Workspace is the ability to send out a link to a project via email (or text, or Slack, etc.) so that other users in your organization do not need to navigate to a project — they can just click the link you send them. This is now possible as well, making it easier than ever to democratize analysis. After saving a project, simply click the “Link to Project” button in the toolbar at the top of the page, and Analysis Workspace will generate a link which you can send to your colleagues.
Note that the recipient of the link you send out does NOT need to have the project shared with him or her prior to clicking through the link. Don’t worry, the recipient won’t be able to make changes to the project unless he or she has complete rights to Analysis Workspace (but can save a new copy, if desired, while leaving the original untouched).
Custom Date Ranges in Date Selector
One of the most underrated features of Analysis Workspace is the ability to build “custom date ranges.” You’ve always been able to select any date range for reporting, but custom date ranges allow you to perform any date comparison you like in data tables and visualizations. For example, Adobe Analytics has a “Last Quarter” date preset, and I can both select that as my date range and use it in data tables for comparing data across dates. With custom date ranges, I can build a date range that Adobe Analytics does not have by default, such as “Two Quarters Ago,” and use that to compare last quarter with the one before it.
The piece that was missing, that many of you have requested, is the ability to use that custom date range (e.g. “Two Quarters Ago”) as a preset for selecting your reporting date range using the calendar module. That is now possible with this release. All of your custom date ranges will automatically show up as calendar presets, making it even easier to select relevant date ranges for your projects.
In this screen shot, the bottom for date range presets are custom ones that I created. When I select them, the dates they represent are automatically selected in the calendar and applied to my project.
By the way, I’ll be sharing some tips for creating powerful custom date ranges in my session at Adobe Summit in both Las Vegas and London (March 20-24 and May 11-12, respectively), so I hope to see you there!
Analysis Workspace has always been fast. It cuts through billions of rows of data to return results to you in a few seconds. We’ve never seen a tool that queries this much data as fast as Analysis Workspace does. And we made it faster.
To be clear, the enhancements that we made to Analysis Workspace in this release do not cause data queries to run any differently. But these changes make the UI work much more smoothly in your web browser, especially if you have a bunch of complex data tables and visualizations in the project you’re working. We noticed that sometimes, when you dive really deep in your analysis, or have a bunch of stacked segments in multiple data tables, dragging and dropping could get sluggish. Those issues have been resolved, making Analysis Workspace even faster than it already was.
There are actually even more new features and enhancements that I didn’t have space to describe in this already-too-long post! Check the Adobe Marketing Cloud release notes for more information.
Much More Coming Soon!
As huge and exciting as these updates are, there is even more coming soon, including full support for scheduled and downloadable reports (in PDF format!), additional analysis tools, new visualization types . . . the list goes on and on. As always, please let us know your thoughts on Analysis Workspace by posting on the Idea Exchange or by tweeting at me. We love hearing from you!